The invention relates to a temperature-dependent nonlinear compensation algorithm for a soil temperature and humidity sensor. The algorithm comprises the following steps of (1) obtaining nonlinear modeling and coefficient estimation according to experimental data, wherein the experimental data includes soil moisture, soil temperature and real soil moisture, and obtaining a table 1, (2) nonlinearlydescribing a volumetric water content Theta by using the dielectric constant epsilon b of water according to an empirical Topps equation, (3) solving the equation by using measurement humidity Thetai, meas of each row in the table 1 and calculating and obtaining epsilon b(i), and (4) solving the equation by using real humidity Theta i, nl of each row in the table 1 and calculating and obtainingepsilon b(i, T). The algorithm has the advantages that a third-order temperature-dependent nonlinear system model is provided, a corresponding nonlinear parameter estimation algorithm is provided, andthe temperature-dependent nonlinear compensation algorithm for the soil temperature and humidity sensor of the invention has the advantages of high precision, universal applicability and fewer parameters.